Dr. Siegelbaum is also a professor of neuroscience and pharmacology and chair of the Department of Neuroscience at Columbia University College of Physicians and Surgeons.
Control of Synaptic Function, and Spatial Learning and Memory, by the Hyperpolarization-Activated HCN Channels
Steven Siegelbaum's laboratory studies how the regulation of synaptic transmission and neuronal integration are important for learning and memory, focusing on the hyperpolarization-activated channels present in neuronal dendrites.
As a high school student, Steven Siegelbaum read a book by Isaac Asimov that would shape his future scientific career. The Human Brain: Its Capacities and Functions introduced Siegelbaum to the idea that neurons can create electrical signals and that, by measuring those signals, one could begin to understand cellular communication in the brain at the level of simple physics. "That made a big impression on me," he says. "It triggered a spark that I seem to keep coming back to."
For much of his career as a professor of neuroscience and pharmacology at Columbia University, Siegelbaum has studied how electrical signals in brain cells are generated and controlled by ion channels. Ion channels are proteins in cell membranes that act as gates on the cell surface, allowing certain charged particles to pass through the membrane. In many of these channels, electrical signals open and close the gates. Siegelbaum is intrigued by how those stimuli, along with other molecular signals called second messengers, can influence the activity of a cell and, in turn, an animal's behavior.
In the late 1990s, Siegelbaum had been working with one kind of ion channel, called the cyclic nucleotide–gated channel, when a postdoc in the lab next door—the lab of HHMI investigator and Nobel laureate Eric Kandel—inadvertently found a protein that resembled the channels Siegelbaum had been studying. Together, the two labs discovered that this protein is indeed an ion channel and is encoded by one of a family of genes known as pacemaker channels. These pacemaker channels were first characterized in heart cells, where they generate the electrical signals that cause the heart to beat spontaneously. However, the channels are also found in the brain, where their function is less clear.
"We had actually, several years earlier, tried to clone these pacemaker channels, based on the idea that they would have a cyclic nucleotide–binding site, but it didn't work. But by serendipity, a postdoctoral scientist, Bina Santoro, who was working next door to us happened to clone these channels by chance." This was the first successful attempt to clone the gene that later became known as HCN1.
With that discovery, Siegelbaum and Kandel set out to determine what these channels do and how they work. They decided to make a knockout mouse, in which the HCN1 gene was deactivated. Unlike the other pacemaker genes, HCN1 is only expressed in brain cells, but because the other pacemakers influence heart rhythms, the researchers suspected that HCN1 might play a role in the rhythms of the brain. But when they knocked out the channels, the brain rhythms of the mice remained relatively unchanged. Surprisingly, though, the mice got smarter. The mice lacking these pacemaker channels could complete a water maze much faster than their normal littermates.
Intrigued by this unexpected discovery, Siegelbaum pressed further. "Why has evolution put a [pacemaker] channel there? It's obviously not there just to make the mice stupid. It's got to be doing something important."
These channels are normally concentrated at the very tips of the dendrites of certain neurons in the hippocampus, a part of the brain that is important for spatial learning and memory. Dendrites are the branches that extend from the body of the cell and receive electrical signals from other neurons. Siegelbaum thinks that the HCN1 channels may be blocking the dendrites from processing signals from other cells and are, therefore, preventing the mice from learning. When the channels are disabled in the mice, there is nothing to stop those signals from being processed in the neuron. He has also found that, without the channels, there is an improvement in long-term potentiation—the process by which long-term memories are believed to be stored—but only in the dendrites that have a high concentration of HCN1 channels. "So our idea is that the channel is acting as an inhibitor of learning and memory," he says.
There is still much work to be done before he can know this for sure. And Siegelbaum is eager to find out what the beneficial purpose of the channels is. "Our favorite hypothesis is that the channel is there to act as a gate of learning and memory, allowing the brain to select what information gets stored by controlling whether the channels are on or off," he says. Maybe when the channels are off all the time, a mouse (or human) can't distinguish between relevant and irrelevant information. Maybe it learns too much.
One thing that Siegelbaum has learned is how many surprises research can bring. "Some of the most interesting results from my lab have happened when we were in the middle of performing an experiment to answer one particular question, but we obtained an unexpected finding that happened to be even more interesting than what we were originally looking at," he says. "That is one of the great pleasures of doing science."